This proposal is an amended competing renewal application to extend our currently funded project, R01 MH056888, Interpersonal Functioning and Emotion in Borderline Personality. Based on our prior results, we continue to regard attachment theory as the best framework for understanding interpersonal functioning in borderline personality disorder (BPD) In BPD, anxious, ambivalent attachment takes a distinctive form- frequent disappointments and anger toward others lead to periods of hostile withdrawal, followed by loneliness and fears regarding separation, producing new attempts at approach and reconciliation, which continue until the next episode of frustration and disappointment. The key themes in this self-defeating cycle are (a) struggles with anger and (b) an inability to maintain a consistent and comfortable interpersonal distance from others. We propose to examine these processes from a dyadic perspective-an innovative, new step for research on personality disorders. We will assess these processes from the perspectives of both the patient and a romantic partner, given our earlier findings that romance is a high-risk domain for patients with BPD, consistent with our assumption about the central role of attachment mechanisms. The major objective is to enhance our understanding of these central mechanisms in BPD by assessing these phenomena at multiple levels of analysis. We will use three methods (at different levels of granularity and with different time scales) to achieve this objective. First, we will conduct a dyadic laboratory interaction involving a discussion of a currently troublesome and unresolved issue in the relationship. The interaction will be coded for both affect and behavior using the Specific Affect Coding System (SPAFF), and measures of peripheral psychophysiology will be recorded. Second, we will use a dyadic, ecological momentary assessment (EMA) protocol completed simultaneously by participant and partner for three weeks. Finally, we will include a social network analysis of the 25 most emotionally significant figures for the participant (and partner) during the past year. Our first specific aim s to document that patients with BPD (n = 55 dyads) have higher mean levels of anger, greater variability in anger, and a stronger association between threats to attachment and anger compared to patients with other PDs (OPD, n = 55 dyads) or no PD (n = 55 dyads). Our second specific aim is to document that patients with BPD display greater variability in interpersonal distance compared to patients with other PDs or no PD. Our third specific aim is to document that these effects regarding anger and interpersonal distance are moderated by the characteristics and behavior of partners, especially their own attachment anxiety. The work will be done with a total sample of 165 dyads consisting of men and women between the ages of 18 and 40. We believe that incorporating dyadic assessment (together with continued attention to the broader social networks in which both partners are embedded) is the best way forward to improved, evidence-based treatments that can address interpersonal functioning at the level of day-to-day transactions.